What is DivX Labs ?

[Gej]

DARC team has a new toy,
You all have a new toy.

DARC team’s has a secret hideout: DivX Labs. Here in our labs, we are free to do whatever we want, including releasing newly prototyped features and interesting codec tools before they’re ready for primetime. Getting everyone’s feedback early in the research and development process will help us make better feature decisions and choose optimal settings or “tweaks” within features. As soon as we have a newly prototyped feature we’ll release it on DivX Labs (and ONLY on DivX Labs). For each new release, we’ll provide information on the purpose of each feature, how to use the feature and possibly some initial test results.
DivX Labs is intended for those with serious existing knowledge of video compression. We will be keeping things very technical here.


Teleport yourself to DivX Labs: http://labs.divx.com/

[Sirber]

Is that a joke?

[anomiERIC]

all your base are belong to us

[dTb]

Originally posted by Sirber
Is that a joke?

No.

[bond]

looks like a big testing zone...

do we get paid for testing? :D

[cordraconis]

Originally posted by bond
looks like a big testing zone...

do we get paid for testing? :D

Yeah, first backstabbing the open source project @ ProjectMayo (oh no, the community hasn't forgot :o ) by going commercial without any further notice, then asking the same community to do free béta-testing and suggestions.

Am I the only one who sees the irony in that?
:D

Disclaimer: Let's call it "Stress from the Exam-period."
:sly:

[Darrius "Junto" Thompson]

Originally posted by cordraconis
Yeah, first backstabbing the open source project @ ProjectMayo (oh no, the community hasn't forgot :o ) by going commercial without any further notice, then asking the same community to do free béta-testing and suggestions.

Am I the only one who sees the irony in that?
:D

Disclaimer: Let's call it "Stress from the Exam-period."
:sly:

Hmmmmmm.. Backstabbing? ProjectMayo was not just about having code that was open it was about sharing data and knowledge. I won't go into detail because I've done it one to many times but yes we stopped our open work at ProjectMayo and continued to focus on what was most important... building a good codec. This was not an easy decision but it came down to us wanting to ensure we could continue to build a good codec, continue advanced video related Research & Dev activities, and build other products and services around codecs. Those that actually did any dev at ProjectMayo are now fulltime employees or have worked with us till this day. It's great we were also able to provide a place they could get payed to do what they loved....Create cool stuff around video. So having source open to closed makes one commerical or is just because we also need to make a living and actually charge for some products?

Can you believe it, we actually have to find ways to make a few dollars so we can support our efforts much of which go towards things those who have continued to stay enthusiastic and support us have asked for... pushing DivX into all sorts of devices, software apps, games (Last one was Enter the Matrix... Woohoo), etc. That's not an easy job but is needed for us to continue to build good stuff. If you think these choices are easy give it a shot. Oh and of course now I can afford that Ford Pinto or Gremlin Escape From Seaworld sort of car I always wanted. Truly a dream come true :eek:

>by going commercial without any further notice, then asking the >same community to do free béta-testing and suggestions.

Uh..... are are you also saying we shouldn't ask for and listen to your suggestions? I think you got the message incorrect about DivX Labs and Beta Testing. Maybe this was our fault. That's why we call it Omega or Prototype testing. Let me try and clarify. Anytime we do a release or even prior to a release there are so many parameters that can be tweeked which affect visual quality technically and subjectively. With just 6 of us here to work on this (and DivX Hardware, Gaming SDK's, etc..) it doesn't provide as good of feedback we could get from the many fresh eyes that can offer suggestions on what settings\parameters work best with the divx codec PRIOR to a full release. Staring at this all day long sometimes makes it harder to see the subjective differences and by the time we get home our eyes are ready to fall out of their sockets. Anyways the feedback is used to help us make sure we release features that offer improvement while using the right parameters. A really cool prototype feature will be ready in several weeks. Can wait for you all to see it. I'll see if I can get some screenshots on the feature soon and post it somewhere. It's sweet! Well I like it and can you imagine that part of the features in this came from "suggestions" :p

Anyways, your thoughts are still appreciated since it helps me to remember that my view of DivX and who we are doesn't always match everyone elses. So in all honesty your frankness is appreciated.. damn you! :o I like to think we're just half a dozen guys trying to provide and build good stuff with the chance to get to do this job for years to come without having to be corporate asshol*s and while making sure DivX doesn't just have our warped brains in the codec but YOURS! Damn balance is hard. If anyone is ever in San Diego, CA you are welcome to contact me and come see for yourselves what a bunch of geeks we really are.

>Disclaimer: Let's call it "Stress from the Exam-period."

I hope the stress increases and remember you have 6 guys at DivX now wishing you bad things for your Exam.. karma :devil: Actually good luck.

Cya at DivX Labs! ;)

Darrius Thompson
Director of myself and R&D DivX

[cordraconis]

OK, it's a bit late to write an answer, but I'll see to it tomorrow. :o

Goodnight!

[cordraconis]

A friend of mine who's going to work for Nvidia to program drivers during the summer, advised me to reply with "whatever" and leave the topic alone. :D
But I'll try to be a bit more friendly than that ;)
(I like to believe my reputation on this Forum is not bad and I'd rather keep it this way. :) )

After re-reading my post (which was more a reaction about the "getting paid for Béta-testing" - hence the Quote (!) - than it was a reaction @ DivXnetworks) and your post, I understand your reaction.

Originally posted by Darrius "Junto" Thompson
Hmmmmmm.. Backstabbing? ProjectMayo was not just about having code that was open it was about sharing data and knowledge. I won't go into detail because I've done it one to many times but yes we stopped our open work at ProjectMayo and continued to focus on what was most important...
<cut>
So having source open to closed makes one commerical or is just because we also need to make a living and actually charge for some products?

I'm not going to go into the details of Economy either, but basically once you start making profit, it's commercial.
(And I'm not believing you if you would say you 6 guy's are only charging for the costs of PC's and Pepsi. :D)
You are right about one thing: If this is your Job, then you obviously want to get paid.


Originally posted by Darrius "Junto" Thompson

<cut>
If you think these choices are easy give it a shot. Oh and of course now I can afford that Ford Pinto or Gremlin Escape From Seaworld sort of car I always wanted. Truly a dream come true :eek:

Absolutely not!
I have 2 uncle's who both have their own buisness. We don't talk that much about it, but I do know it's a jump in the dark and taking a risk of starting your own company. I'll never say it's an easy choise.
(BTW, I googled for that Ford Pinto, and it's a very old car so I don't think it costs that much. :P I couldn't found anything about Seaworld and Gremlin, Escape or both these last words.
Whatever. http://www.vtk.ugent.be/forums/images/smiles/nixweiss.gif )

Originally posted by Darrius "Junto" Thompson

Uh..... are are you also saying we shouldn't ask for and listen to your suggestions?
<cut some explanation about divX Labs.>

OK, so if I'm correct it's about parameters.
My opinion in this would be to let the users themselves decide which numbers to fill in the boxes. On the other hand, not everybody is willing to experiment on this, and the "right-from-the-box" users want of course the best possible defaultsettings to use in their game/program/... So somebody has to smooth the edges. I would have asked the béta-testers to do this, or if you really wanted more eyes, then I'd have asked Doom9 (I assume he's a béta-tester.) to release it on the forum and ask the testers who are more familiar with these settings to give their judgements.
Oh well ...

Originally posted by Darrius "Junto" Thompson

I hope the stress increases and remember you have 6 guys at DivX now wishing you bad things for your Exam.. karma :devil: Actually good luck.

Thanks!
(And if it's a comfort for you: I lost a few hours of studying by googeling the net, and reading the old Project Mayo forums etc ... ;) And even then I'm sure I missed a lot of stuff.
You guy's better make all this trouble up with the next release of DivX ... :sly:

Greetz,

Cordraconis

[vinouz]

Well I'm very glad this version is out. For my own part I'm always happy when a new codec comes and it's time to test it.
If you wanna make profit, god bless you.
If the codec proves to please me, if it responds to a need or satisfies me, then I take it. And this includes the terms of use. I understand your need. For my part I love to test, tweak, encoding real films doesn't interest me, as as usual I rent the dvd, think about copying it, try some encodes, then scraps them or keep sequences of them for further encoding testing. I see a film once or twice, and the if I'm gonna see it twice, it's at friends or a new place, or here with re renting the dvd if there are new friends a majority of which haven't seen it, and I don't have the need for keeping a copy of a film once I've seen it. So I rent the DVDs, for 1.5€ (1.6$) it's perfect to me.
And for the codec, I use the adware, which if bothers me (and it didn't really happen) I deactivate (Ctrl-Alt-Sup, regedit and start folder).
Here in tahanea, I don't even have any adware (just looked at it, it seems clean. And I know my computer so I could almost put my hand to fire, [well, not, I didn't nmap it ATM, [and won't - there are plenty of other things more interesting to do with my time : the sun is shining ! BUt well, let's finish quickly this post, after which I g o to the "Calanques" (french marseille cliffs to the sea. I wish you went to them, to understand what nmapping my comp would mean in my order of priorities now) ;p ]]). So it totally corresponds me. I help tahanea, don't get bored, people wanting to use divx pro without effort to make easy encodes agree to give DARC money, and hardware manufacturerd build players on which they can see their CDs. They're perfect, having CDs of their holiday video or whatever [what they put, they assume - i assume my short clips for tweaking encoding for my understanding and testing purposes, I'm glad with that] and they can see them on their sofa. They agree to pay, meaning they accept the deal.

I'm glad you have your ford pinto, I didn't know myself, and I'd be glad if I build as big and good a project you built, make it a successful product, and be able to have what I dream of : a boat [with a friend, we have a deal : the first one who can offer him the boat pays it, the other one puts all the concurrent money he gained, buy food, drinks and takes cash. And turn around the world, Here We Are !].

But it's not the point. The point is that I am glad with this new baby, and will test it.

Thanks guys !

Vince

[Darrius "Junto" Thompson]

Originally posted by cordraconis

I'm not going to go into the details of Economy either, but basically once you start making profit, it's commercial.
(And I'm not believing you if you would say you 6 guy's are only charging for the costs of PC's and Pepsi. :D)
You are right about one thing: If this is your Job, then you obviously want to get paid.

Agreed I would then consider this commercial. So then the point I was trying to make is projectmayo was actually also commercial from the start. We had every intention of making money to support ourselves and our newly formed company. So we were commercial from almost the beginning when we got our first dollar by selling some really bad t-shirts we had created. They were really bad. Anways agreed and yes the cash we make goes to alot more than PC'S and Pepsi, since today is friday we are getting Pizza, Beer, and a few hours of Unreal Tournament.. Actually I think today we are playing Wolfenstein so of course we had to buy a few copies of these games too :D. But of course there are other initiatives we are undertaking that require funds. Also keep in mind that sometimes when a commercial venture is started the goal behind it may be more philosophical rather than a goal driven toward money although the money is sometimes needed toward accomplishing a goal that is about creating what are hopefully good changes.

(BTW, I googled for that Ford Pinto, and it's a very old car so I don't think it costs that much. :P I couldn't found anything about Seaworld and Gremlin, Escape or both these last words.
Whatever. http://www.vtk.ugent.be/forums/images/smiles/nixweiss.gif )

That's funny you looked it up. Yeah the ford pinto is a disgusting car so disgusting I might want one some day. The Gremlin is like the AMC Pacer which is the ugly car in "Waynes World" and I call it the "Escape from Seaworld" car because it looks like it could be a good aquarium because of the big windows.

http://wayne.s.world.free.fr/Wayne/garthMobile.html

OK, so if I'm correct it's about parameters.
My opinion in this would be to let the users themselves decide which numbers to fill in the boxes. On the other hand, not everybody is willing to experiment on this.

Yes it's definitely alot about picking parameters. Actually this is coming soon with some of the new features. We will expose all the parameters for tweaking for those who are willing to try them out. For example Psychovisual modeling has several parameters that we really had to determine for each level and these are not exposed in the GUI but you'll see these and other parameters soon.

Thanks!
(And if it's a comfort for you: I lost a few hours of studying by googeling the net, and reading the old Project Mayo forums etc ... ;) And even then I'm sure I missed a lot of stuff.
You guy's better make all this trouble up with the next release of DivX ... :sly:

:) Hope the exams go well. Back to pizza, beer, unreal... no wonder we have bugs in our code....

Thanks cordraconis

Cya..
Darrius Thompson
DivX Corporate Bastard :eek:

[Darrius "Junto" Thompson]

Originally posted by vinouz

I'm glad you have your ford pinto, I didn't know myself, and I'd be glad if I build as big and good a project you built, make it a successful product, and be able to have what I dream of : a boat [with a friend, we have a deal : the first one who can offer him the boat pays it, the other one puts all the concurrent money he gained, buy food, drinks and takes cash. And turn around the world, Here We Are !].

But it's not the point. The point is that I am glad with this new baby, and will test it.

Thanks guys !

Vince

Thanks Vince appreciate your reply! See Cordraconis there are some great people that actually can enjoy what we are doing. :p Sorry for taking this forum topic somewhat off course.

Darrius

[Acaila]

Sorry for taking this forum topic somewhat off course.I personally don't mind at all. The fact that you're reading/posting here and the news that you've decided to put new features out for public betatesting tells me that the DARC team is very concerned with the users of the DivX codec.
I hope this doesn't come off as too negative, but it's been my experience that the DivX developers hardly ever post anywhere (that is, here or divx.com), so I'm always glad to see you people around. I know you're busy developing a codec, but some active involvement in the way of participating in forum discussions to me seems essential in getting in touch with your userbase.

Which is why I was very positively surprised to see a public betatesting in the form of DivX Tahanea being presented. Although I admit that public testing can have its downsides as well, I believe this does allow people to 1) have the opportunity to get a closer peek into the inner working of the codec, and 2) give the general public the feeling that they can in some way aid/suggest future development.

Anyway, hope to see you around here some more.

[BlackSun]

Nice to see you there Darrius, it's been a long time since we discussed. In the past that was Dan 'BetaBoy' Marlin (who works with me on CoreCodec) who was spreading the DivX speech and he did it very well and I'm very glad that you do the job as good as he did.

Since BetaBoy left DXN, the corp was hidden being a shield of silent but you guys broke it and it is certainly appreciated by a lot of people (like myself).

So keep up the good work...

[Ewi]

But perhaps the main picture on labs.divx.com should be changed... It shows a professor who has put some cables into our monkeybrains and is watching what the monkey is doing with his invention...

Mmmmhhhh...

[Darrius "Junto" Thompson]

Originally posted by Ewi
But perhaps the main picture on labs.divx.com should be changed... It shows a professor who has put some cables into our monkeybrains and is watching what the monkey is doing with his invention...

Mmmmhhhh...

Well it's suppose to be funny and it is much better than having the cables being plugged directly into the opposite end of the monkey with him sitting on his head!:eek: And we could do this too if you'd like :) Uh but your right maybe the message comes across wrong! Maybe we should put DARC badges on the monkey?

Darrius

P.S. Good to see you too BlackSun

[mf]

Junto! :) When are we going to have that multiplayer match between sNL and DXN? Of course you don't remember :D.

The reasons behind the shutdown of OpenDivX and creations of DivX4 are well understandable. Everybody knew the license. But the pulling of encore2 was still a bit, well, "strange". Not the best move imho, especially because it was done in secrecy.

On to a more on-topic subject, I don't get DivX5's features. There's a QPel which I have heard doesn't help much (it should keep sharp sources sharp - to summarize it easily), only h.263 quant which doesn't help DivX's already bad sharpness, there's a completely useless GMC (ok so XviD's GMC is useless too- but a 3-warp-point version will be out soon), and only 1 consecutive B-frame with fixed quantizer ratio. DivX already isn't really a winner in encoding comparisons, and then you're limiting its performance in these ways. Any reasons? I don't think it'd be time limitation, the real hardcore coders in XviD that contribute stuff regularly are only a few - I think 5 or 6. And they do it in free time. Stability or compatibility also wouldn't be an issue - if you avoid nonspec features like modulated quant, XviD is actually more compliant than DivX, strictly speaking.
And now there's DivX Tahanea, which has some form of R-D. Why, if DivX (afaik) has no inter4v? Isn't it a better idea to implement that first? Or the other things I was talking about? Sometimes I really don't get the decisions made. :-/

Old time PMayo forum member :rolleyes:, mf :D

[Darrius "Junto" Thompson]

Originally posted by mf
Junto! :) When are we going to have that multiplayer match between sNL and DXN? Of course you don't remember :D.

mf! Of course I remember... I remember planning how my redeemer was going to find its way to you :D We still play every friday uh and sometimes alot more. I recall the issue was our time difference in playing since we usually play
here on Fridays starting around 4 or 5 Pacific time. We're ready at this time if you can be :)

The reasons behind the shutdown of OpenDivX and creations of DivX4 are well understandable. Everybody knew the license. But the pulling of encore2 was still a bit, well, "strange". Not the best move imho, especially because it was done in secrecy.
[quote]
For the record, I'm not even sure or agree that the decision we made was the right or wrong decision. This remains to be seen. Yeah we pulled encore2 quickly and unfortunately I can't discuss all the reasons we did this but I can say our goal stayed the same.. improve the DivX Codec while getting it into other places besides basic video software (gaming, consumer electronics, etc..). At the time there were some restrictions placed on us (DivX) where we might not have been able to do this if we did not make these decisions where this goal may have been much harder or where companies may have refused to work with us. Looking back I still cannot say if this was the best choice since that it is still an unknown and may stay that way... I'm not sure there was any benefit to the decision we made but what is done is done. But again the thing we still focus is the codec and other technology to push it into other areas while making it easy and to use and more importantly trying to be open about our decisions while communicationg openly when possible and when we have time! mf you always have to stir things up! :) :devil:
[quote]
On to a more on-topic subject, I don't get DivX5's features. There's a QPel which I have heard doesn't help much (it should keep sharp sources sharp - to summarize it easily), only h.263 quant which doesn't help DivX's already bad sharpness, there's a completely useless GMC (ok so XviD's GMC is useless too- but a 3-warp-point version will be out soon), and only 1 consecutive B-frame with fixed quantizer ratio. DivX already isn't really a winner in encoding comparisons, and then you're limiting its performance in these ways. Any reasons?

Yeah quite a few reasons :) Read my threads on the the Hardware Section of the Doom9 forum. http://forum.doom9.org/showthread.php?s=&threadid=55458 . The decision we make also have to take into consideration of what can be supported in hardware with reasonable cost and within a reasonable timeframe. We could require Qpel and GMC and keep trying to improve it however supporting it in hardware makes the issue more than just adding technical complexity... cost and availability of hardware that can perform are the major issues.

The quality improvement that GMC and Qpel give are likely not worth the cost increase in hardware at this time. We have to focus on what can be used on a wide variety of platforms. So it's also a quality vs complexity vs cost issue. Would Qpel and GMC be worth a huge additional cost in the DVD\DivX\XviD\MP4 player that you would buy, maybe it would be but you also have to realize the guys building the hardware also have to take a risk in spending the extra money to support these features which increase their risk in building and selling these products. So we've lately tried to focus mostly on new features that improve quality while not adding complexity to scalability into other devices. You'll see more and more of these features in DivXLabs. We know where the codec is weak and where it is strong and we are attacking the weak areas now. So each time we look at a feature we also have to think about other devices such as hardware. Could we support various quantizer matrices and adaptive b-frames? Sure, but at what cost to hardware? It's not just the increase in cost to you but the risk the manufacturer is willing to take for the cost of such features. Remember then if we add a feature that might make their new chip break this would be a very bad user experience not to mentione upset our parnters!. We need to focus on things that improve quality while keeping compatiblity for the near future so the experience is good and where you don't have half of your files inoperable with your hardware player. Some of the new features the DARC team is working on are going to improve the any areas the codec may be weak and then some. And of course this improvement will be subjective but so far so good.

I don't think it'd be time limitation, the real hardcore coders in XviD that contribute stuff regularly are only a few - I think 5 or 6. And they do it in free time. Stability or compatibility also wouldn't be an issue - if you avoid nonspec features like modulated quant, XviD is actually more compliant than DivX, strictly speaking.
And now there's DivX Tahanea, which has some form of R-D. Why, if DivX (afaik) has no inter4v? Isn't it a better idea to implement that first? Or the other things I was talking about? Sometimes I really don't get the decisions made. :-/

Actually we did have some time limitations since as mentioned we have been working on a few things (mostly hardware support) as well as the core codec. We finally have a few video hardware experts that have helped to share the tasks so we can now get back to finally focusing on the core codec. Our roadmap kind of went like this... DivX Core Codec (Software)-->Hardware-->Back to Core Codec and a few pit stops along the way like Gaming (Enter The Matrix... that was cool and many more games are coming we're helping with), security devices (the cams you see in cop cars!), medical devices (yepp medical.. recording surgeries), and a whole lot more, it's everywhere! So next time you get in a car chase and end up in the hopspital you'll automatically have the whole thing in DivX... remember to wear clean underware :eek: We got hit hard by many requests to get DivX into many places. We got DivX into those places and then got ourselves some more help to take on some of the things our small team did on top of just basic codec development. What a rollercoaster! Anyways we are trying our best now to ONLY focus on core codec work thus as soon as we got help we released DivX Labs so we could start cranking out new and frequent releases.

Anyways we're working on it and with a vengeance! Oh and in regards to XviD having 5 or 6 active developers working on the core code... it's probably less. It's about 3 and communicate quite often offering each other ideas and help. It's a great relationship, the XviD team (Christoph, Pete, and Michael) are REALLY smart guys, as I've learned from the relationship we've built between our 2 teams. They will definitely help all of us create good technology and not just video technology. So at least you have 2 choices from 2 smart teams both trying to do good things! I'm sure the XviD team can vouch for the complexity and time involved in all the things it has taken to get to where we are today. They face similar things have we've discussed and helped each other in dealing with some of these battles. It's a good thing for the all of us (the entire community) since there are our 2 teams working toward the same goal of creating good codecs and other technologies. 2 small, smart and focused teams are much better at dealing and competing with the Mammoths that would like us to go away :)

Why can't I just stop with a 1 or 2 sentence answer? Sorry! Ok now I actually have to get back to work. Cya mf!

Darrius

[mf]

Aah! Well I'll see if I can get something to happen, even if I have to organize a 3-day-no-sleep lanparty for it! :devil:

So, the reasons for the partial support of MPEG4's features is hardware. Hrm. Understandable. However, with XviD simply doing what it wants (whacking in features at random - you should hear sysKin some times :D it's quite a joke; "I have this great idea for <insert feature here>"-- 2 days later: "wee, 0.2dB PSNR increase!"), and its userbase growing steadily, they also demand hardware support. So far most of the hardware players have been crap; not even supporting DivX fully, but WHEN there are finally DivX Certified players, they still won't support XviD. And then those users will be dissatisfied about that, and somewhere along the road players created that are not DivX Certified, but do support the full ASP as XviD creates it. I don't know if DXN gets any royalties from the certification program, but you'd be missing out on it. Plus a label "will play XviD" - even partially(!), might wipe out the trustworthiness of the "DivX Certified" label. Of course this is further on in the future, as you say hardware manufacturers are still struggling to fully support DivX5, but still. There's the possibility. I've seen people from Asia on this forum asking for guidelines on hardware player creation, with replies like "should fully support XviD". You don't know who might be crazy enough out there to actually go and do that.
Anyway, with all this hardware and compatibility stuff, I know what I'll do if I want to play MPEG-4 on TV (not that I want to, a 19" at 1280x960 is MUCH bettter than a 50Hz TV of any size, and the 100Hz ones still have low res, not mentioning their price), I'll go out and get a micro-ATX system (like shuttle), and whack in an all-in-wonder, a network card and a 100GB drive. Instant DVR as well :).

[cordraconis]

Originally posted by mf
I'll go out and get a micro-ATX system (like shuttle), and whack in an all-in-wonder, a network card and a 100GB drive. Instant DVR as well :).
Indeed, with all these versions of all kinds of codecs running around, one might wonder why they don't sell small PC's with upgradable software. ;)
The firmware upgrades are a good step in this direction, but if I'm not mistaken, then you are still limited by the processing power (or memory to process peak bitrates, as Junto already explained) or the instruction set of the decoding chip. So if a codec would need specific instructions, then even new firmware wouldn't help and you need a new chip, right?

And this brings me to another thing: I remember somewhere reading that MPEG Audio Layer 3 (ie MP3) was only a format, and the full meaning of that never hit me until I read that it only discribes how to Decode the MP3-stream, and *not* how you actually make it. (That was given as the reason MP3 can still be improved in terms of quality, like LAME.)

So, ... If MPEG-4 is also a format, then how come you need firmware upgrades after all? I thought it's just about decoding MPEG-4, regardless of DivX or Xvid ?
Am I missing something here?

[manono]

Anyway, with all this hardware and compatibility stuff, I know what I'll do if I want to play MPEG-4 on TV (not that I want to, a 19" at 1280x960 is MUCH bettter than a 50Hz TV of any size

Nonsense.

[Darrius "Junto" Thompson]

Originally posted by mf
Aah! Well I'll see if I can get something to happen, even if I have to organize a 3-day-no-sleep lanparty for it! :devil:

Of course this is further on in the future, as you say hardware manufacturers are still struggling to fully support DivX5, but still. There's the possibility. I've seen people from Asia on this forum asking for guidelines on hardware player creation, with replies like "should fully support XviD". You don't know who might be crazy enough out there to actually go and do that.

I would love for there to be a solution that can support all the advanced MP4 features plus more but the fact of the matter is there is not a single DSP or ASIC available that can do this and there will not be for sometime unless you have quite a bit of money to spend. So although you may have some people in Asia thinking about how to do this the truth is there is no hardware. Believe me, we're working with all the top chip manufacturers also in Asia. Sure you might get a company to say they have a chip or device that can playback all the advanced features just as you can say a 486 can playback all these features :) Believe me we will try and get as many good features in these chips but again we are limited by the current state of technology which will improve in time.

Anyway, with all this hardware and compatibility stuff, I know what I'll do if I want to play MPEG-4 on TV (not that I want to, a 19" at 1280x960 is MUCH bettter than a 50Hz TV of any size, and the 100Hz ones still have low res, not mentioning their price), I'll go out and get a micro-ATX system (like shuttle), and whack in an all-in-wonder, a network card and a 100GB drive. Instant DVR as well :).
Yeah that was my option until we got one of the first KISS DP-500 prototype boxes in here. It does so much more and looks great. It really impressed me when I saw the lastest one and convinced me of how great this type of product is. I love the network connectivity. We were playing files directly on our TV, via the KISS DP-500, that were being progressively streamed\downloaded from the internet in realtime without having a PC anywhere nearby making a ton of noise and taking up space next to the TV. Once I experienced it I realized how powerful these devices will be. Although it is not yet certfied it is close and if you have the opportunity to try one out I would advise doing so. I know I'm going to buy one or sneak the final sample prototypes we will test into my car :)

Darrius "Junto" Thompson
Director R&D DivX

[mf]

Originally posted by manono
Nonsense.
Sorry for having more sensitive eyes than you do. My 19" runs at 99Hz when running in 1280x960.

[manono]

We may be talking apples and oranges here, but here goes. You didn't say if yours was an LCD monitor with a DVI connection or a CRT on a VGA connection. I'll assume for the sake of argument that you have a CRT monitor. And I'm in NTSC land.

As you may know, there are plenty of HDTVs and projectors with DVI connections now. I've got an HDTV with native 1280x720 resolution that can scale to 1080i. It also has DVI in. I've also got a DVD/MPEG4 player with DVI out that can send 480p, 720p or 1080i. Digital to digital. You sit in front of your monitor watching an analog picture. I sit on my sofa watching big screen all digital. You figure out who has the better viewing experience.

[mf]

Originally posted by manono
I'll assume for the sake of argument that you have a CRT monitor. And I'm in NTSC land.
My mention about 99Hz would indicate a CRT, yes. And I don't use TFTs because they can't switch resolution properly (fixed physical resolution). And I'm in PAL land (my mention about 50Hz would indicate that).

As you may know, there are plenty of HDTVs and projectors with DVI connections now. I've got an HDTV with native 1280x720 resolution that can scale to 1080i. It also has DVI in. I've also got a DVD/MPEG4 player with DVI out that can send 480p, 720p or 1080i. Digital to digital. You sit in front of your monitor watching an analog picture. I sit on my sofa watching big screen all digital. You figure out who has the better viewing experience.
1) In PAL land we don't have HDTV (at least not in .nl), nor any HDTV sets
2) You quoted me on "a 19" at 1280x960 is MUCH bettter than a 50Hz TV of any size". That would mean you were replying to my statement that a 19" CRT Monitor (there are no 19" TFTs that run at 1280x960) is better than an analogue (hence the 50Hz) CRT television. First off, 50Hz TVs always run at broadcast resolution, which is 768x576 in PAL land (overscanned so actual viewable rectangle is about 704x540), and they're 50Hz. 50Hz, even if it's interlaced, hurts my eyes. So refreshrate-wise and resolution-wise, I prefer my 19".
3) I have looked at many 100Hz (digital?) TVs, of which lots looked very interesting (and made a huge contrast in quality to the little 14" 50Hz one that was hanging from the ceiling next to them, I must say) - especially the Philips ones which I believe had framerate doubling. However, the resolution was with all impaired to PAL resolution, which didn't look very nice.
In NTSC land, where HDTV is available, I would agree with you. But here it just isn't.
And even then, your argument doesn't stand, since I still had never mentioned anything about HDTV sets, and was only comparing to the TV types available to me. So it is indeed apples and oranges (or should I say apples and mangos? I bet if HDTV were a fruit it'd taste lots better than regular TV :D).

[dTb]

@mf, have you checked out any tv's with vga input? While it still requires you to have your pc in the lounge I'm very happy with the quality of my tv. I'm also in a PAL territory and it's a widescreen samsung plano that I own, the vga input uses 60Hz progressive scan and the picture is akin to a monitors. All in all I much prefer it to my 19" monitor (although your monitor sounds a lot better than mine :)).

[manono]

Hi mf-

My mention about 99Hz would indicate a CRT, yes.

I thought as much, but I'm not knowledgeable enough to have been sure. My CRT computer monitor will only do 85Hz at that resolution.

And I'm in PAL land (my mention about 50Hz would indicate that).

That much I knew, hence my comment about me being in an NTSC country.

..since I still had never mentioned anything about HDTV sets,

That was the reason for the apples to oranges comment.

In any event, this is a bit off topic. Doom9 is soliciting comments about MPEG4 players, so maybe later I'll post in the Player Forum about my experiences with mine.

And I'm also happy that Junto has started to visit. His comments, while long, are very informative, and I look forward to reading and learning more.